Cooperative supervision of livestreaming e-commerce based on stochastic evolutionary game and overconfidence

Abstract

The rapid development of livestreaming e-commerce has been accompanied by an increasing number of misleading marketing behaviors (MIBs) that require adequate regulations. To reveal the impact of irrational behavior (overconfidence) and to explore effective supervision strategies tailored to different types of streamers. We have classified streamers into brand-affiliated streamers and professional streamers (including internet celebrity streamers and ordinary streamers), then four types overconfidence are identified and integrated into a stochastic evolutionary game framework to investigate the regulatory effectiveness. The findings indicate that platform overconfidence positively affects supervision, while overconfidence among streamers and consumers has the opposite effect. For brand-affiliate streamers, the reputation mechanism exerts the most significant regulatory influence and should be heightened; additionally, enhancing platform penalties proves effective in cases of streamer overconfidence, whereas reducing supervision costs works better in other scenarios. Regarding professional streamers, a combination of platform penalties and incentive mechanisms leads to more stable and effective regulatory outcomes, especially in cases of streamer or consumer overconfidence. Furthermore, for internet celebrity streamers, the reputation mechanism serves as a beneficial supplement; for ordinary streamers, reducing regulatory costs proves to be more effective. Therefore, this study provides insights for classified and tiered regulation policy formulation regarding livestreaming e-commerce and provides a new perspective for supervision research by integrating overconfidence and evolutionary games.